introduction to strategic interactions, duopolies and ... 12, number 2, 2015/2... · model to...
TRANSCRIPT
Australasian Journal of Economics Education
Volume 12, Number 2, 2015, pp.12-29
INTRODUCTION TO STRATEGIC
INTERACTIONS, DUOPOLIES AND COLLUSION:
A CLASSROOM EXPERIMENT*
Julien Picault
University of British Columbia, Okanagan
ABSTRACT
This classroom experiment introduces strategic interactions and collusion in an
environment where two producers are interacting in a market. It introduces the
three classic duopolies: Bertrand, Cournot and Stackelberg, and compares the
effects of competition versus collusion. The experiment is intended for use prior
to presentation of the theoretical models, as it allows students to develop their
understanding of duopolies during the game and then to compare this
understanding with the theory. The use of this experiment allows the instructor to
present the intuition behind strategic interaction in an attractive manner and to
dissociate intuition from mathematical complexity.
Keywords: Experiment, Bertrand, Cournot, Stackelberg, Collusion.
JEL classifications: A22.
1. INTRODUCTION
Strategic interaction is a key concept that economics and business
students must master in various microeconomics and industrial
organization courses. When presenting strategic interaction, there is a
significant increase in difficulty for students. Duopolies are typically
an economics student’s introduction to strategic interaction and this
topic often follows chapters on perfect competition and monopoly.
When presenting monopoly and perfect competition, the mathematical * Correspondence: Julien Picault, The University of British Columbia - Okanagan,
3333 University Way, Kelowna, BC, Canada, V1V 1V7. Tel: (+1) 250-807-9227; E-
mail: [email protected]. I am grateful to Faithe Reimer and classroom participants
in the Industrial Organization and Regulation courses at UBC in Winter 2011, Fall
2012, Fall 2013 and Fall 2014. Thanks also to two anonymous referees.
ISSN 1448-4498 © 2015 Australasian Journal of Economics Education
Strategic Interactions, Duopolies and Collusion 13
resolution and intuition of both are straightforward. However, the
mathematical resolution of duopolies is often confusing for students,
especially in the Cournot and Stackelberg cases.1 Students need to
acquire mathematical skills while gaining an understanding of the
model to effectively select the right path of resolution. Indeed, it is the
first time that they are required to use intuition to solve a model. At
this point, students experience an increase in the complexity of the
material which can decrease their motivation. Therefore, the chapter
on duopolies is a pinnacle opportunity to encourage student interest in
the material.
Simulating markets in class provides an attractive way for students
to develop their knowledge and understanding. Holt (1999) explains
that “one of the most exciting recent developments in the teaching of
economics is the increased use of classroom exercises that insert
students directly into the economic environments being studied”.
Furthermore, the literature indicates improved achievement among
students taught using experiments. Frank (1997) compared learning
outcomes of a specific concept, for both students who participated in a
classroom experiment and those who did not; he concluded that the
participating students were more successful. Dickie (2006) and
Durham, McKinnon & Schulman (2007) found similar results.
Durham, McKinnon & Schulman (2007) also show that classroom
experiments positively affect students’ retention of economic
knowledge.
The experiment presented in this paper is helpful as it introduces
students to duopolies through an interactive environment, where they
can begin to acquire economic intuition behind strategic interactions
in duopolies without being overwhelmed by the mathematics. This
experience helps simplify the understanding of the mathematical
resolution of the models. A key feature of the experiment is that it
allows the separation of intuition and mathematical resolution in the
learning process. Using this experiment addresses the two learning
objectives in sequence instead of simultaneously.
In the mathematics education literature, it has been demonstrated
that learning highly abstract concepts is facilitated by tangible
applications.2 This explains the growing interest in developing new
in-class approaches to mathematics which require students’ active
1 Formal presentation of duopolies can be found in Cabral (2000).
2 See Giroud (2011) for more thorough details about this literature.
14 J. Picault
participation through hands-on activities. Bartolini Bussi (2009)
discusses the benefits of using experiments in the learning of abstract
mathematical concepts. In her paper, she observed that if the result of
a problem is provided too early, students’ attention is focused on the
final result rather than on the dynamic process of resolution. This
reinforces the need to experiment with concepts before providing
students with the mathematical resolution. Jahnke (2007) also
suggests that results coming from empirical evidence, such as an
experimental approach, can convince students more than rigorous
proofs.
Additionally, Becker & Watts (1995), Durham, McKinnon &
Schulman (2007) and Emerson & Hazlett (2012) indicate a connection
between the use of experiments and improved student engagement.
Classroom experiments can improve engagement because students see
the experiment as a game rather than a lecture. Regardless, while
“playing the game” they are being exposed to an economic situation
from which they can develop intuition and understanding of the
models to be presented in class.
There is currently an important debate in the literature on the use of
paper-based vs computerised experiments in economics education.
Computerised versions of oligopoly markets are available on different
on-line platforms and provide a decent alternative to paper-based
experiment. One benefit of computerized versions is the reduced
amount of administration for the instructor. Carter & Emerson (2012)
investigated both methods and their results suggest that student
achievement is comparable in both cases; however, students tend to
view in-class experiments more favourably than computerized
experiments. In other words, using paper-based experiments
effectively reinforces engagement and motivation.
Earlier articles such as Meister (1999), Beckman (2003), Ortmann
(2003) and Giles & Voola (2006) focus on Bertrand and/or Cournot
duopolies and/or involve some degree of mathematical complexity.
The game presented in this paper is designed to deal with a broader
set of duopolies and can easily be adapted to multiple contexts. It
offers a flexible learning environment, free of mathematical
complexities. This is particularly helpful in introductory courses
including: Principles of Microeconomics; Introduction to Industrial
Organization; and Managerial Economics. It may also serve as a
review for intermediate level courses.
Strategic Interactions, Duopolies and Collusion 15
During this experiment, groups of students are divided to play the
roles of different firms. They determine how firms make production
and pricing decisions in an environment where they and their
competitor are influencing market conditions. This creates an
opportunity for students to discuss strategies in teams, stimulating
understanding and learning. Students can compare the different
implications of the potential structures of various duopolies such as
Bertrand, Cournot, Stackelberg and Collusion.
Brauer & Delemeester (2001) identify a potential risk to using
classroom experiments in that students may not play the game
seriously and could, therefore, devalue the use of such experiments.
To avoid this, students’ performance during the game will account for
1 percent of their final grade.3 Students are also required to write an
individual assignment for an additional 1.5 percent of the final grade.
In total, therefore, the game determines 2.5 percent of students’ final
grade. Furthermore, it is possible to give immediate reward for firms
earning the highest profits during the experiment. A piece of candy for
every treatment winner and a gift card for the team having the largest
profits overall can serve as immediate incentives. Small reward can
create a friendly competition and makes the students even more
invested in the experiment.
In the sixth chapter of Holt (2007), it is mentioned that when
players are able to recall the history of play, which they can in this
experiment, chances of tacit collusion are increased. I did not
experience much of this problem while running this experiment. My
understanding is that students understand that each treatment is played
just a few times and they usually focus on short-term profits. An
instructor, who is concerned about tacit collusion, can rematch firms
after each period to severely decrease the likelihood of tacit collusion.
Although it is difficult to measure the precise impact of this
experiment on the comprehension of economics students,
improvements have been observed. More specifically, in Fall 2014
and Fall 2015, I collected data on the ability of students to utilize the
correct method of resolution to solve a Cournot duopoly and a
Stackelberg duopoly during the midterm exam following the
experiment. Table 1 presents this data. In Fall 2014 and Fall 2015,
27% of the students did not participate to the experiment. This gives
an opportunity to compare results of students who did and did not
3 See below for assignment details.
16 J. Picault
Table 1: Success Rates when Solving Duopolies
Success Rate in Solving
the Cournot Duopoly
Success Rate in Solving
the Stackelberg Duopoly
Attended the experiment 89% 89%
Did not attend the
experiment 73% 50%
participate in the experiment. While vertically comparing success rates
of students who attended and who did not attend would be purely
informative,4 we can learn from comparing horizontally success rates
between Cournot and Stackelberg duopolies since they require the
same ability. Students that attended the experiment had a success rate
of 89% in both the exercises about Cournot and Stackelberg
duopolies, while students who did not attend had a success rate of
73% for Cournot and 50% for Stackelberg. This comparison suggests
that students who participated were more able to differentiate and
solve accurately the Cournot and Stackelberg mathematical
resolutions.
The remainder of the paper details the following: Section 2 provides
a description of how to include this experiment in a lesson plan;
Section 3 describes the experiment; Section 4 presents the
mathematical resolution of the models; Section 5 describes some
advantages associated with this experiment; and Section 6 concludes.
2. LESSON PLANNING
This experiment has been used multiple times in an Industrial
Organization and Regulation course, where the study of duopolies is
done after monopoly. Before taking this course, students normally
have completed Principles of Economics courses. In my Industrial
Organization course, the study of duopolies typically requires three
lectures for a total of four and a half hours of classroom time.5 During
the first lecture the experiment is conducted. After completing the
experiment, students are required to prepare an individual assignment
to be submitted at the beginning of the second lecture. This
4 One can argue that the distribution of ability is not the same when considering
students attending and not attending the experiment. 5 Note that I do not discuss collusion in the same chapter as duopolies.
Strategic Interactions, Duopolies and Collusion 17
assignment requires students to amalgamate the intuition developed
during the experiment. The assignment consists of a two page report
guided by the following questions:
- For each treatment, describe how your strategy evolved. What
would you do differently now?
- How did you assess your opponent? Did the concept sharing of
demand influence your decision making?
- For every treatment, should the strategy of each firm be the
same. Why or why not? Explain your observations.
The first two sets of questions are designed to illustrate the main
intuitions while the last question is to verify the students’
understanding of the difference between simultaneous and sequential
games.6 The assignment also assists the instructor in detecting
potential misunderstandings of the proper intuitions. It provides an
opportunity to discuss directly with students their personal challenges
and to identify gaps in their learning prior to the exam.
The second lecture is used to present theoretical models and to
create a discussion of the experiment based on the assignment
questions. The third lecture is then devoted to exercises which include
the mathematical resolution of duopolies. Data from the experiment
can be used to create these exercises. Students’ confidence improves
when they realize the solution of both the experiment and the
mathematical resolution are the same.
3. THE CLASSROOM EXPERIMENT
For this experiment, the class is separated into groups (firms) of two
to five students.7 To keep the experiment manageable, use no more
than eight firms playing four markets. For larger classes, teaching
assistants can support the management of additional markets. Each
firm receives an instruction sheet and firm ID comprised of a letter
and a number.8 The letter represents the market in which they will be
competing and the number represents the firm. For example, A1
represents firm 1 in market A; and A2, firm 2 in market A. The only
numbers used are 1 and 2 since the game is created for duopolies,
however it is possible to create multiple markets if necessary. Placing 6 In simultaneous games, firms should have the same strategy, while, in sequential
games, they should not. 7 Typically, I ask students to form themselves groups before distributing the
instructions of the experiment. 8 See the instruction sheet in the Appendix.
18 J. Picault
all firms designated with 1 on the left side of the classroom and all
firms designated with a 2 on the right side prevents unauthorized
discussion.
Before starting the experiment, read the instructions to the students
and present them with the sheet used to record the results (these are
provided in the Appendix). Allow for questions to ensure that students
understand all the instructions and how to record the results. To
remove any mathematical complexity, the demand schedule takes the
form of a table in the instruction sheet.9 It is important to emphasize
that firms share the same market demand with their competitor. The
experiment consists of four treatments corresponding to the Bertrand,
Cournot, Stackelberg and Collusion cases, and it is a good idea to
review the instructions specific to the treatment each time it changes.
Each treatment includes numerous market periods and represents a
specific market structure. Depending on the time allowed for the
experiment, the number of periods within each treatment may vary. It
is best to play at least two periods in each treatment and to stop
playing when markets have converged to the equilibrium. This
typically takes two to four market periods depending on the treatment.
A specific treatment may be targeted to the outline of the class as
they are played independently. To keep the class interested, decisions
made by each firm can be displayed on the board. Also, quick market
periods are preferred. It is helpful to have an online timer projected on
the screen in the classroom to keep track of the length of the periods.10
Also, when preparing the experiment, the instructor may want to
adjust the record sheet in order to predetermine the number of market
periods to be played.11
Table 2 shows the results from one particular
experiment I conducted in class.
(a) Bertrand
The first treatment is a simulation of a Bertrand competition between
two firms. The firms have to simultaneously determine the price at
which they sell a product on the market. The firm with the lowest
price wins the entire market. If both firms price the same way, they
will have to share the market half and half. The initial period for
Bertrand is limited to three minutes, with subsequent periods of two
9 This is also provided in the Appendix. This is similar to the demand schedule used in
Holt (2007), chapter 6, and Badasyan et al. (2009). 10
Many free timers are available online. 11
See record sheet in appendix.
Strategic Interactions, Duopolies and Collusion 19
Table 2: Results from a Classroom Experiment
Period Treatment # Q1 Q2 Price
Cost
per
Unit
Profit 1 Profit 2
1 1 10 0 18 (20)12
4 140 0
2 1 0 14 12(14)12
4 0 112
3 1 11 11 6 4 22 22
4 2 10 14 4 4 0 0
5 2 8 10 10 4 48 60
6 2 8 8 12 4 64 64
7 3 22 2 4 4 0 0
8 3 14 6 8 4 56 24
9 3 12 6 10 4 72 36
10 4 6 6 16 4 72 72
11 4 6 10 12 4 48 80
12 4 8 10 10 4 48 60
minutes each. Firms should use this time to determine the price. By
the end of the period, they must have a price written on a small piece
of paper with their firm ID. The instructor collects the papers and
displays the prices on the board. Firms, then have enough information
to calculate profits acquired during the period. Displaying profits on
the board ensures information is the same across firms and students
are documenting accurate calculations on their record sheet. Once
record sheets are completed, another period can begin and periods are
repeated until the end of the treatment. This treatment is illustrated by
the first three periods in Table 2.
(b) Cournot
The second treatment is a simulation of a Cournot duopoly. Firms
simultaneously determine the quantity of a product that they will offer
on the market. The total quantities produced by firms for the market
are tallied to determine the price using the demand schedule outlined
in the game instructions. For Cournot, the initial period is three
minutes with subsequent periods of two minutes each. Firms must use
12
The number in parentheses corresponds to the price proposed by the firm that does
not supply in a Bertrand competition if any.
20 J. Picault
this time to determine the quantity of product that they will supply on
the market. At the end of the period, they must write the quantity on a
small piece of paper with their firm ID. The instructor again collects
these papers and displays quantities and prices on the board. The firms
then have enough information to calculate profits acquired during the
period. Once record sheets are completed, periods are repeated until
the end of the treatment. This treatment is illustrated by periods 4 to 6
in Table 2.
(c) Stackelberg
The third treatment is a simulation of a Stackelberg duopoly. Firms
sequentially determine the quantity of a product that they will offer on
the market. Stackelberg involves a different timeline. Firm 1
determines quantity first. Then, firm 2 observes firm 1’s quantity and
determines its own quantity of the product based on this information.
For this treatment, each firm should be provided two minutes to
establish their quantities. Firm 1 uses this time to determine the
quantity of product that they will supply on the market. At the end of
this time, firm 1 must have a quantity written on a small piece of
paper with their ID. The instructor collects their papers and transfers
the quantities to the board. Once firm 2 can see the quantity, the time
begins. At the end of this time, firm 2 must have a quantity down with
their firm ID. The instructor once again collects these and displays the
quantities on the board. The firms then have enough information to
calculate profits accumulated during the period. Once record sheets
are completed, the next period can begin. This treatment is illustrated
by periods 7 to 9 in Table 2.
(d) Collusion
To ensure order during this treatment, firms should choose a
“diplomat” before beginning. The role of the diplomat is to negotiate
an agreement with the other firm. Once each firm chooses, the fourth
treatment can start. The fourth treatment is a simulation of a Cournot
duopoly where firms may decide on a common strategy. Firms
strategically determine the quantity of a product that they will offer on
the market. For this treatment, a period of two minutes is given for
firms to decide on the proposal they make to the other firm. Then,
diplomats meet in the centre of the classroom to negotiate an
agreement for two minutes. Then, another two minutes is given to
discuss within the firm the quantity to produce. Firms determine
Strategic Interactions, Duopolies and Collusion 21
whether they honour their agreement. At the end of the period, all
firms must have a quantity written on a small piece of paper with their
firm ID. The instructor will take all the pieces of paper and document
the quantities on the board. Firms, then have enough information to
compute profits collected during the period. Once record sheets are
completed, another period can begin; this is repeated until the end of
the treatment. This treatment is illustrated by periods 10 to 12 in Table
2.
4. RESOLUTION OF THE MODELS
Assume a market where two symmetric firms are supplying the same
good. The inverse demand function for the good is p = 28 - Q and the
total cost for firm i is iii qqTC 4)( for i =1, 2. Each market simulated
in the experiment is based on this demand function which provides a
good basis for comparison.
(a) Bertrand
The Bertrand equilibrium predicts that when firms are symmetric, the
price equals marginal cost. Therefore, using the data on which the
experiment is based, p = 4 + ε where ε represents the smallest
monetary denomination. In the context of this experiment ε = 2,13
and
students should end up with a price of p = 6.
(b) Cournot
Given our context, profits of the firms in a Cournot competition are:
11211 4)28( qqqq
22212 4)28( qqqq
Profit maximization yields the best response functions for both firms:
2/12 21 qq
2/12 12 qq
We can then derive Cournot quantities:
81 q
82 q
Inserting these quantities in the inverse demand yields the price on the
market as p = 12 and the profit of each firm will be 6421 .
13
See the market demand schedule in the Appendix.
22 J. Picault
During this treatment, it may take students a few periods to accurately
identify the correct quantities.
(c) Stackelberg
The timeline of the Stackelberg duopoly played in this experiment is
that firm 1 chooses first and firm 2 second. We begin with the best
response function of the second firm that was calculated in the
Cournot treatment:
2/12 12 qq
We next insert the value of 2q in firm 1’s profit function, yielding:
2/12 2111 qq
Firm 1’s profit maximization leads to the value of 1q . Inserting this
value for 1q into the best response function of the second firm yields
2q . So, the Stackelberg quantities are:
121 q
62 q
Inserting these quantities into the inverse demand function yields the
price in the market as p = 10. Firm 1 will make a profit of 721 and
firm 2 will make a profit of 362 . It should be remembered that it
may take students a few periods to accurately identify the correct
quantities.
(d) Collusion
When students are required to negotiate an agreement with the
competitor, they must weigh the benefits of establishing an agreement
against competing in a Cournot duopoly ).64( 21 If an
agreement is strategically negotiated, each firm will produce 50% of
the monopoly quantity and will, therefore, split monopoly profit
50/50. In this market, a monopoly would have the following profit
function:
QQQ 4)28(1
Profit maximization yields the following values:
12Q
16p
144
Strategic Interactions, Duopolies and Collusion 23
If the firms share the monopoly profit, each firm obtains a profit of 72.
When comparing the potential profits, it appears that collusion is more
beneficial than Cournot. However, firms have the option to renege on
their agreement. Firms should use their best response function and the
information that the other firm is producing half of the monopoly
quantity, to determine whether to honour the agreement. Without loss
of generality, firm 1’s decision can be used as an example to
determine the best strategy. Firm 1’s best response function is
2/12 21 qq , and if firm 1 is convinced that firm 2 will be
producing 62 q , then firm 1 should renege on the agreement by
producing 91 q . This would result in firm 1 making a profit of 81.
The theory predicts Cournot should be played at every round in a
finite environment. To maintain collusion, an indefinitely repeated
game and a credible threat are necessary. Students will usually collude
as long as neither firm disregards the agreement. If a competitor has
reneged on their promise, students typically produce the Cournot
quantity in the subsequent period as illustrated in the last row of Table
2. Since I have never been able to play this treatment more than 3
consecutive times, it is hard to make strong conclusions about the type
of strategy students are using. However, I have never seen students
colluding after being reneged. This tends to reflect a grim strategy.
5. PEDAGOGICAL ADVANTAGES
As explained in the introduction, many pedagogical advantages come
with the experiment. Also, this experiment can be used to reach
additional pedagogical goals. Usually, differentiating between
Cournot and Stackelberg mathematical resolutions is one of the most
significant difficulties students have. These resolutions are very
similar and therefore can be a source of confusion. The experiment
improves students’ understanding of the mathematical resolutions. It
provides them the foundational reasoning and dynamic intuition
required to mathematically solve duopolies. Therefore, instructors can
refer to the experiment when explaining the different steps. This
analogy helps the student to link the mathematical steps with their
concrete experience of the steps.
Students commonly ask “how do we know when to solve the
system of the best response functions; or when to insert firm 2’s best
response function into firm 1’s profit function?” This question refers
to the selection of the steps needed to solve Stackelberg and Cournot
24 J. Picault
duopolies. During the experiment, students have already experienced
the correct steps. This improves their understanding of the differences
between simultaneous and sequential decisions, therefore providing
them the ability to use intuition to select the correct resolution. Having
the ability to build on this experience simplifies the learning and the
instructor ability to clearly answer this question.
Another commonly asked question is “why should firm 2’s best
response function be inserted in firm 1’s profit function?” This
question refers to the Stackelberg duopoly. Prior to determining the
quantity to be produced, the firm that moves first must predict how the
second firm will respond. This is precisely what students have
experienced during the game. When presenting the Stackelberg
duopoly in class, it is helpful to ask a student from a firm 1 to explain
how his/her firm determined the quantity during the game. This
reminds the class of the right intuition in the students own words and
provides an experiential explanation. Highlighting the link between
the resolution and the experiment facilitates students’ understanding.
Solving in class the models presented in section 3 improves
students’ confidence in finding the resolutions. The experiment has
been designed so the practical results will correspond with the
mathematical results. In other words, students will find the same
numbers when mathematically solving the model as were found
during the game. This validates the model for students and increases
their confidence in the material.
Finally, this experiment provides intuition useful for the chapter on
collusion. Typically, at least one group will renege on the agreement
which introduces students to the limitations of collusion. Students
already know what collusion is, and they usually see collusion as
something easy to maintain. The experiment will challenge this belief.
It is also a simple way to introduce repeated games with a
predetermined endpoint. During the collusion treatment, students
usually will set up and maintain a collusion until a firm discovers it is
in the firm’s best interest to break the agreement. This is similar to the
way we demonstrate that collusion is inappropriate with a
predetermined endpoint. At least one player will not benefit from
collaborating in the last period and therefore the agreement will not be
maintained in the penultimate period and repeating this pattern back to
the first period.
Strategic Interactions, Duopolies and Collusion 25
6. CONCLUSION
This in-class experiment is simple for the instructor as it does not
require much material because it is not a computer-based experiment.
It also does not consume much class time as it is designed to last
between 90 and 120 minutes depending on the time that is available.
The paper also explains how this experiment can be introduced in a
lesson plan for the study of strategic interactions through duopolies. It
is especially designed for intuition assimilation and it creates the time
for students to understand the dynamical process before receiving the
solution. Playing this experiment also offers munitions to the
instructor to answer the most common questions that students have by
referring to their experience during the game. This experiment
provides students an environment free of mathematics where they can
experience and understand strategic interactions and duopolies. It is an
asset when teaching these concepts as it particularly assists students
later on when students are required to understand steps of the
mathematical resolutions.
This experiment is also designed to allow for clear comparability of
the outcomes of the three common forms of duopolies, namely
Cournot, Bertrand and Stackelberg, as well as Collusion. This is
particularly helpful later when theoretically demonstrating these
differences; the instructor can refer to students’ experience to justify
the pertinence of the theory.
APPENDIX: IN-CLASS GAME INSTRCUTIONS FOR STUDENTS
Overview of the Experiment
This experiment is intended to expose you to strategic interactions,
duopolies and collusion. You will be expected to explain the process used
to determine your actions and describe the steps of your reasoning.
Each team will be assigned the role of a firm. As a firm, you will sell your
products to customers. Firms will be matched with one another at the start
of a game and will remain matched with this firm throughout the game.
Each firm will be assigned a number either 1 or 2 allowing you to compete
in a duopoly market.
Do not forget, your goal is to maximize your firm’s profits!
26 J. Picault
Your Firm
Your firm is producing the same good as your competitor. The cost of
producing each unit is 4$ and there is no fixed cost associated with the
production of the good.
Your profit or loss is evaluated by calculating the difference between the
price and your unit cost multiplied by the number of units your firm sells to
customers.
For example, if the price is $12 and your quantity sold is 6 units, your
profits would be:
(Price – Unit Cost) x Quantity = ($12 - $4) x 6 = $48
Your Competitor
Your competitor is producing the good for the exact same cost as your
firm.
Market Demand
The demand for the good is provided in the following demand schedule:
Price ($) 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0
Qty 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
When each period is concluded, please record the price, the quantity
(number of unit sold) and your profit, on the record sheet provided.
Your Role
Your role is to make strategic decision for your firm to maximize profit. Do
not forget that your competitor also influences the market.
The experiment consists of 4 treatments of multiple market periods. The
market structure will change in each treatment.
First Treatment:
At the beginning of each period, each firm must choose its price for the
product. If your price is higher than your competitor’s price, you will not
sell any goods. If your price is lower than your competitor’s price, your
goods will supply the whole quantity demanded at this price. If your prices
are equal, you will share the market 50/50 (each supplying half of the
quantity demanded). At the conclusion of each period, please calculate
your profit.
Second Treatment:
At the beginning of each period, each firm must choose its quantity to be
sold. You must combine your quantity with the quantity of your competitor
Strategic Interactions, Duopolies and Collusion 27
and find it in the demand schedule to determine the price on the market. At
the conclusion of each period, please calculate your profit.
Third Treatment:
At the beginning of each period, firm 1 chooses a quantity. Firm 2 chooses
its quantity based on the quantity produced by firm 1. You must combine
your quantity with the quantity of your competitor and find it in the
demand schedule to determine the price on the market. At the conclusion of
each period, please calculate your profit.
Fourth Treatment:
Before beginning this sequence each team must choose a “diplomat”. The
role of the diplomat will be to negotiate an agreement with the other firm.
At the beginning of each period, diplomats can negotiate an agreement in
term of quantities to be produced. Then, each firm must choose its quantity
to be sold. If you have an agreement with your partner, you can decide to
either honor or disregard the agreement. You must combine your quantity
with the quantity of your competitor and find it in the demand schedule to
determine the price on the market. At the conclusion of each period, please
calculate your profit.
Market Information
After each period, the price, the quantities sold in the market and the profits
for each of the parties will be displayed on the board.
Any Questions?
Record Sheet Firm Number: Teammates:
Period Treatment # Your
Quantity
Quantity
of your
Competitor
Price Cost per
Unit Profit
1 1 4
2
4
3
4
4
4
5
4
6
4
7
4
8
4
28 J. Picault
REFERENCES
Badasyan, N., Goeree, J. K., Hartmann, M., Holt, C., Morgan, J.,
Rosenblat, T., Servatka, M. and Yandell, D. (2009) “Vertical Integration
of Successive Monopolists: A Classroom Experiment”, Perspectives on
Economic Education Research, 5, pp.61–78.
Bartolini Bussi, M. G. (2009) “Experimental Mathematics and the
Teaching and Learning of Proof”, CERME 6 Proceedings, pp.221-230.
Becker, W. E. and Watts, M. (1995) “Teaching Tools: Teaching Methods
in Undergraduate Economics”, Economic Inquiry, 33 (4), pp.692-700.
Beckman, S.R. (2003) “Cournot and Bertrand Games”, Journal of
Economic Education, 34 (1), pp.27-35.
Brauer, J. and Delemeester, G. (2001) “Games Economists Play: A Survey
of Non-computerized Classroom-Games for College Economics”,
Journal of Economic Surveys, 15 (2), pp.221–236.
Cabral, L.M.B. (2000) Introduction to Industrial Organization, Cambridge,
Mass., and London: MIT Press.
Carter, L. and Emerson, T. (2012) “In-class vs Online Experiments: Is
There a Difference?”, Journal of Economic Education, 43 (1), pp.4-18.
Dickie, M. (2006) “Do Classroom Experiments Increase Learning in
Introductory Microeconomics?”, Journal of Economic Education, 37 (3),
pp.267-288.
Durham, Y., McKinnon, T. and Schulman, C. (2007) “Classroom
Experiments: Not Just Fun and Games”, Economic Inquiry, 45 (1),
pp.162-178.
Emerson, T. and Hazlett, D. (2012) “Classroom Experiments: International
Handbook on Teaching and Learning Economics”, Hoyt, G. and
McGoldrick, K. (eds), International Handbook on Teaching and
Learning Economics, Cheltenham, U.K. and Northampton, MA, USA:
Edward Elgar, pp.90–98.
Frank, B. (1997) “The Impact of Classroom Experiments on the Learning
of Economics: An Empirical Investigation”, Economic Inquiry, 35 (4),
pp.763-769.
Giles, M. and Voola, J. (2006) “The Shrimp Game: Engaging Students in
the Classroom”, Australasian Journal of Economics Education, 3 (1/2),
pp.73-89.
Giroud, N. (2011) Etude de la démarche expérimentale dans les situations
de recherche pour la classe, Unpubished Doctoral Dissertation,
Université de Grenoble.
Holt, C.A. (1999) “Teaching Economics with Classroom Experiments: A
Symposium”, Southern Economic Journal, 65 (3), pp.603-610.
Holt, C. A. (2007) Markets, Games and Strategic Behavior, New York:
Pearson Addison Wesley.
Strategic Interactions, Duopolies and Collusion 29
Jahnke, H. N. (2007) “Proofs and Hypotheses”, ZDM: The International
Journal on Mathematics Education, 39 (1-2), pp.79-86.
Meister, J.P. (1999) “Oligopoly: An In-Class Economic Game”, Journal of
Economic Education, 30 (4), pp.383-391.
Ortmann, A. (2003) “Bertrand Price Undercutting: A Brief Classroom
Demonstration”, Journal of Economic Education, 34 (1), pp.21-26.